Most probable first-round upsets in the men's 2026 NCAA tournament

"Only five double-digit seeds won in the first round of the 2025 men's NCAA tournament, and two were 10-seeds, which don't qualify as Giant Killers for the purposes of this column. The Elite Eight was tied for the chalkiest of all time -- all top-three seeds or better -- and the Final Four was just the second ever to feature four 1-seeds."

"NCAA tournament betting lines are larger than ever, and our Giant Killer model doesn't give any team seeded 11th or worse even a 40% chance of winning in the first round. It gives only one No. 12 seed -- the perennial upset pick so many like to identify -- even a 20% chance to bust brackets."

"North Carolina's ceiling is significantly lower with star freshman Caleb Wilson (broken thumb) out for the season. Predictably, the Tar Heels' efficiency numbers dropped off in the eight games Wilson missed, especially on the offensive glass and in 2-point field goal percentage on offense and defense. That opens the door for a VCU team that has won 16 of its past 17 games entering the tournament."

The 2025 men's NCAA tournament has been notably chalk-heavy, with minimal Cinderella stories compared to historical norms. Only five double-digit seeds won in the first round, and the Elite Eight featured exclusively top-three seeds or better, tying for the chalkiest outcome ever. The Final Four included four 1-seeds, only the second time this has occurred. Despite March Madness's appeal being partly driven by upsets, current conditions make them unlikely. NCAA tournament betting lines are at record levels, and predictive models show no team seeded 11th or worse has a 40% chance of winning their first-round matchup. Only one 12-seed receives even a 20% upset probability, compared to last year's four matchups exceeding 40% upset chances. Analysts recommend examining specific high-probability upset scenarios, such as VCU against North Carolina, whose star freshman Caleb Wilson is sidelined with a broken thumb.

Read at ESPN.com